Endoplasmic Reticulum Stress Signaling Comprises a G9a Inhibitor Tolerance Pathway and PERK Inhibition Increases Anti-Leukemia Activity of G9a Inhibitor in Leukemia Cells and Leukemia Stem-like Cells

Background: Histone methyltransferase (HMTase) G9a regulates the transcription of multiple genes by primarily catalyzing dimethylation of histone H3 lysine 9 (H3K9me2), as well as several non-histone lysine sites. Recently, pharmacological and genetic targeting of the G9a was shown to be efficient in slowing down acute myeloid leukemia (AML) cell proliferation in a mouse model and human AML cell lines thus making this HMTase potential target for epigenetic therapy of AML. Activation of adaptive mechanisms to drug plays a crucial role in drug resistance and relapse by allowing cell survival under stressful conditions. Therefore, inhibition of the adaptive response is considered as a prospective therapeutic strategy. The tolerance mechanism to HMTase regulation in leukemia cell is unclear yet. The PERK-eIF2α phosphorylation pathway is an important arm of the unfolded protein response (UPR), which is induced under conditions of endoplasmic reticulum (ER) stress. Recent previous studies showed that pro-survival ER stress is induced in cancer cells and contributes to development of drug resistance.

Methods: We investigated the levels of apoptosis and ER stress by G9a inhibitor BIX-01294 in leukemia cell lines. U937, cytarabine-resistant U937 (U937/AR) and KG1 were used. U937/AR cell line was established in our laboratory by exposing parental U937 cells to stepwise increasing concentrations of cytarabine.

Results: We initially examined the expression of G9a in leukemia cell lines and the primary AML cells obtained from a patient at the different time point. In U937/AR cells and primary AML cells obtained at relapse, G9a expression was increased compare to that in U937 cells and primary AML cells obtained at diagnosis, respectively. G9a expression was also increased in KG1 cells. In both of U937 and U937/AR, apoptotic cell death was induced by BIX-01294 in a dose-dependent manner. In contrast, apoptotic cell death was minimal in KG1 cells which are enriched in cells expressing a leukemia stem cell phenotype (CD34⁺CD38^-). To address the activation of ER stress response by BIX-01294 in leukemia cells, we examined the effect of BIX-01294 treatment on PERK and eIF2α protein expression and phosphorylation levels. We found that treatment of U937, U937/AR, KG1 cells with 3μM of BIX-01294 for 24h caused an upregulation of phosphorylated PERK and eIF2α. The upregulation of PERK phosphorylation was associated with a decrease in PERK protein levels after treatment. To further address the role of the PERK-eIF2α phosphorylation in BIX-01294 sensitivity, we examined whether PERK inhibition using small interfering RNA (siRNA) or specific inhibitor could sensitize cells to BIX-01294-mediated death. The siRNA against PERK effectively inhibited BIX-01294-mediated phosphorylation of PERK and eIF2α in U937 and U937/AR cells. The addition of PERK siRNA led to a significant increase in the extent of BIX-01294-induced apoptotic cell death in U937 (P = 0.0003) and U937/AR (P < 0.0001) as compared with that of BIX-01294 treatment alone. PERK inhibitor GSK260641 significantly increased BIX-01294-induced apoptotic cell death in U937 (P < 0.0001) and U937/AR (P = 0.006) cells. To our surprise, addition of PERK siRNA or GSK260641 increased the sensitivity of KG1 cells to BIX-01294-mediated death in a dose-dependent manner (P = 0.0003 for siRNA, P = 0.0053 for GSK260641).

Conclusion: These data demonstrated that PERK-eIF2α activation has a pro-survival function to G9a inhibitor in leukemia cells and mediates resistance of AML stem cells to G9a inhibitor treatment. The PERK-eIF2α phosphorylation arm may represent a suitable target for combating resistance to G9a inhibitor in AML. The mechanisms underlying the increased sensitivity of AML cells with PERK inhibition to G9a inhibitor are unclear at present and are needed to define in further studies.